quantify the representation of major active faults in all three spatial dimensions. Further objectives should include the following:
the integration of other information, such as three-dimensional seismic velocities and attenuation parameters, surface topography, surface geology, and subsurface geologic horizons, into unified structural representations of active fault systems;
the use of these three-dimensional representations as the basis for combining all available data on geodetic velocities and fault slip rates into kinematically consistent models of deformation zones; and
the extension of these kinematical representations to fully dynamical models that incorporate realistic rheologies, boundary tractions, and body forces. The latter can be inferred by combining surface topography and gravity with density variations measured at the surface and inferred from seismic tomography.
A dynamical description of fault system behavior must include the state of stress and how it changes with time. Stress can be measured directly in the near-surface environment accessible by mining and drilling, but the principal way to infer the stress field at depth is through modeling. The four observational facilities of the EarthScope program will provide critical data for this purpose, including precise measurements of surface deformation gradients and their temporal evolution (PBO and InSAR), detailed images of subsurface structures that control stress and strain heterogeneity (USArray), and better knowledge of deformation processes at depth (the San Andreas Fault Observatory at Depth [SAFOD]).
The research objectives outlined above will require (1) the ability to manipulate large data sets from geology, geodesy, and seismology; (2) the development of novel techniques for analyzing and interpreting these data sets; and (3) substantial information technology (IT) resources for developing, verifying, and maintaining community models and making them available to a heterogeneous, widely distributed group of users. The considerable experience accumulated by the petroleum industry over the last decade has shown that the construction of faithful and flexible structural representations is a difficult task, requiring significant resources and substantial reliance on advanced IT tools.
The considerations detailed in Section 5.3 highlight the importance of determining the composition, structure, and physical state of fault-zone materials and damaged border zones. At larger scales there is a need for